提出基于最小二乘支持向量机(LSSVM)的弓网系统广义动态建模方法,并对弓网模型进行辨识及仿真研究。首先对离散化后的弓网系统动力学模型的线性部分和非线性部分进行分析,构建ARX Hammerstein弓网系统模型,然后根据LSSVM的方法对模型参数进行辨识,最后利用弓网实际检测数据对该弓网模型进行弓网接触力的建模预报试验。试验结果表明,LSSVM-ARX Hammerstein弓网模型可以有效地逼近弓网系统接触力的动态行为,具有较高的预报精度,从而验证了所提方法用于弓网系统广义动态建模的有效性。 A generalized dynamic modeling method for bow network system based on least squares support vector machine (LSSVM) is proposed. The bow mesh model is identified and simulated. Firstly, the linear part and non-linear part of the discretized bow system dynamics model are analyzed to build the ARX Hammerstein bow system model, and then the LSSVM method is used to identify the model parameters. Finally, Bow - web model for bow mesh contact force modeling prediction test. The experimental results show that the LSSVM-ARX Hammerstein arch mesh model can effectively approximate the dynamic behavior of the bow mesh system with high prediction accuracy and validate the effectiveness of the proposed method for the general dynamic modeling of arch mesh systems.